蛋白質的動態關連性

Abstract

有些蛋白質內部的氨基酸殘基並沒有直接和配體（ligand）有鍵結, 但是這些殘基卻能巨幅地影響蛋白質和配體間的鍵結親和力(十倍或更高). 這些氨基酸可以以丙氨酸掃描式突變在蛋白質的內部交界處來偵測得, 稱為熱點氨基酸(hot spot residues). 為了讓實驗進行地更有效率, 我們用我們發展出的方法發現蛋白質的鍵結位和熱點氨基酸間在幾何結構上具有高度的相關性, 可以辨識出這些熱點氨基酸和非熱點氨基酸. 分子模擬可以複雜的分子勢能來分析蛋白質內部的分子運動模式, 是一套非常有力的工具, 但是它非常地耗時. 因此, 我們使用簡化的物理分析模式, 如彈性網路模式(GNM)、中心模式（CM）、權重原子接觸模式(WCN)來探討蛋白質內部的動態關連性. 然而, GNM的交錯關連矩陣不能辨識出熱點氨基酸. 改良過的WCN和常態化後的CM可以辨識出這些熱點氨基酸. 我們發展出的這些方法可以顯示出熱點氨基酸和鍵結氨基酸間在幾何關係上具有高度地相關性. 最後, 也發現鍵結氨基酸和高度保留的氨基酸間也有高度的空間相關性.

There are residues in proteins not directly bind to ligand, but they can dramatically affect the binding affinity(tenfold or higher)between protein and ligand. These residues called hot spot residues which can be detected by alanine scanning mutagenesis in protein interface. For the purpose to design experiment effectively, we develop some methods to find the geometric correlation between binding sites and hot spot residues.These methods can identify hot spot residues and non-hot spot residues. The molecular dynamics(MD) simulations with complex molecular potential energy can analize the motion of proteins. MD simulation is a powerful tool, but it is time-consuming. For this reason, we use simplified mechanical methods such as Gaussian network model(GNM), and our methods, centroid model(CM) and weighted contact-number model(WCN) to discuss the dynamical correlation in proteins. Howerer, the cross-correlation matrix in GNM can not identify the hot spot residues. The improved WCN and normalized centroid model can identify these hot spots. The methods we developed can show that there are high correlation between functional hot spot residues and binding sites in geometrical relation. In the end, we can also find the high spatial correlation between binding sites and highly sequence conserved residues.